One of the most pressing issues in modern vaccine development is the poor heterosubtypic neutralizing antibody responses elicited by highly diverse viruses that pose major threats to public health. Elicitation of these so-called broadly neutralizing antibodies (bnAbs) ? or antibodies that bind and neutralize many different viral subtypes ? are rare in human patients, which complicates vaccine development and allows for repeated annual viral infections (e.g., Influenza and Dengue) or even superinfection (HIV) in human patients. In particular, current vaccines elicit bnAbs at insufficient titers for long-lasting protection against all currently circulating and pandemic Influenza strains. Therefore, it is critical to understand why some Abs are selected over others (i.e., why they are immunodominant), an issue directly relevant to the design of effective vaccines for many recalcitrant infectious diseases. Our long-term objective is to map the development of antibody lineages in a way that enables the prediction of likely responses as a tool to manipulate the process of antibody selection. Our primary objective is to test the hypothesis that the antibody evolutionary landscape is more limited for anti-Influenza bnAbs than for subtype- specific Abs, with this difference explaining why these bnAbs are not often selected at high levels in humans. To test this hypothesis we will display Influenza-specific antibody libraries on yeast and use a transformative sorting and deep sequencing pipeline to evaluate each member variant of the library for its affinity and nonspecific binding properties. This massive functional data will then be used for a network analysis to reconstruct plausible evolutionary trajectories for each somatic bnAb and subtype-specific head Ab. The motivation for the proposed research is guided by the urgent need for develop methods to map and manipulate rules of in vivo antibody affinity maturation to develop vaccines against refractory pathogens of high interest to public health including Influenza, Dengue, and HIV. The proposed research project will be carried out by pursuing three specific aims: 1) Determine ontogeny from germ line to mature human antibodies for two heterosubtypic HA stem binders; 2) Determine ontogeny from germ line to mature human antibodies for four heterosubtypic and subtype- specific HA head binders; 3) Determine the number of evolutionary trajectories from a representative germline Ab. This approach is innovative because it combines a unique hypothesis with state of the art protein engineering tools needed to evaluate the hypothesis, and it is significant because the data generated here will illuminate why long-lasting bnAb responses to HA immunogens are so rare. The approach raised in this application may also expedite rational structure-based design of vaccines, prophylactics, and therapeutics against a range of human pathogens.